摘要
目的建立基于径向基神经网络的宫颈癌肿瘤标志物的诊断系统。方法采集体检中心及妇科门诊、病房病例资料。统计血清肿瘤标志物水平,建立径向基神经网络的诊断系统。结果五种血清肿瘤标志物各组间有极显著差异,p<0.01;神经网络的的判别的准确率为100%。结论基于径向基神经网络的血清肿瘤标志物的诊断系统方法可行,性能良好。
Objective To establish the RBF neural network based on the cervical cancer tumor markers in the diagnosis system. Methods To collect the medical center and gynecological clinic, ward ease information. Calculated the levels of tumor markers in serum, and establishment of radial basis function neural network diagnosis system. Results The five kinds of serum tumor markers among the groups had significant difference, p 〈 0.01, and neural network identification accuracy rate was 100%. Conclusion RBF neural network based on the serum tumor markers in the diagnosis system was feasible, with good performance.
出处
《现代医院》
2012年第10期12-15,共4页
Modern Hospitals
基金
广东省科技厅基金资助项目(编号:2009B030801353)
